Why PDF conversions lose formatting
A PDF looks like a finished document, so it's natural to expect a converter to turn it back into one. But a PDF doesn't store a document the way Word does — it stores a picture of the page, drawn with text. Understanding that difference explains almost every conversion problem you'll see.
What's inside a PDF
PDF was created by Adobe in the early 1990s and is now an ISO standard. Its goal is simple: a page should look exactly the same on every screen and printer. To guarantee that, a PDF describes each page as a set of drawing instructions, roughly:
- use this font, at this size, in this color;
- draw these characters starting at this exact position;
- draw a line from here to there;
- place this image in this rectangle.
A Word document works the other way around. It stores structure — paragraphs, headings, lists, tables, styles — and lets the software decide where each line breaks on the page. Converting a PDF to Word means guessing the structure from the finished drawing. That guessing is where things go wrong.
Paragraphs and line breaks
In most PDFs, each line of text is placed separately. There's usually nothing saying "these five lines belong to the same paragraph." A converter has to infer it from spacing and alignment. When it gets it wrong, you get a hard line break at the end of every line — the text looks fine until you try to edit it and nothing reflows.
Even spaces between words aren't always real characters. Some PDFs position each word (or each letter) individually, and the converter has to decide whether a gap is a space, a tab or just tight kerning. That's why converted text sometimes has missing spaces or words glued together.
Columns and reading order
On a two-column page, the PDF may store text in the order it was drawn, which isn't necessarily the order you read it. A converter that reads strictly left to right, top to bottom will interleave lines from both columns. Better converters detect the columns first, but newsletters, brochures and scientific papers with sidebars and captions often still come out scrambled.
Tables
To a PDF, a table is just text positioned in a grid, plus some lines drawn around it. There's usually no notion of rows, columns or cells. A converter has to spot the grid from the alignment of text and the positions of lines. When cells contain wrapped text, merged cells or no borders at all, that becomes very hard, and the table may come out as tab-separated text, as separate text boxes, or as an image.
Tip: if all you need is the data in a table, selecting and copying it straight from the PDF into a spreadsheet often works better than converting the whole document.
Fonts and special characters
- Embedded fonts: a PDF often includes only the characters it uses from each font (a "subset"). If you don't have that font installed, Word substitutes another one and line lengths change.
- Ligatures: combinations like "fi" or "fl" may be stored as a single glyph. A converter that can't map it back to two letters produces odd characters or missing letters.
- Missing character maps: some PDFs don't record which letter each glyph represents. The page looks perfect, but copied or converted text comes out as gibberish. In that case, OCR is often the only way to get usable text.
Headers, footers, hyphens and images
- Headers and footers are repeated on every page and end up in the middle of your text, between paragraphs.
- Hyphenated words split across lines ("infor-mation") stay split after conversion.
- Images and charts may be exported as pictures, split into pieces, or dropped entirely by text-only converters.
- Scanned pages contain no text at all — only a photo of text — so text-based converters return nothing.
Tagged PDFs convert better
Some PDFs include tags: an invisible structure layer that marks headings, paragraphs, lists and tables, mainly so screen readers can read the document properly. PDFs exported from Word, Google Docs or LibreOffice with accessibility options turned on usually have them. When tags are present, converters have much more to work with, and the result is noticeably cleaner. If you create PDFs yourself, see how to export them properly in how to save a Word document as PDF.
How to get the cleanest result
- Ask for the source file. If someone sent you a PDF of a Word document, asking for the .docx is faster than any conversion.
- Match the tool to the document: see our comparison of conversion methods.
- Convert only what you need. Extract the relevant pages first so there's less to clean up.
- Start from a reflowed version. Our Enhanced conversion joins lines into paragraphs and detects headings, which removes most of the tedious cleanup.
- Rebuild tables instead of repairing them. It's usually quicker to insert a new table and paste the data.
Try a reflowed conversion. Get flowing paragraphs and detected headings, processed privately in your browser.
Convert PDF to Word